rpart: Binary Classification

# nolint start
library(mlexperiments)

See https://github.com/kapsner/mlexperiments/blob/main/R/learner_rpart.R for implementation details.

Preprocessing

Import and Prepare Data

library(mlbench)
data("PimaIndiansDiabetes2")
dataset <- PimaIndiansDiabetes2 |>
  data.table::as.data.table() |>
  na.omit()

feature_cols <- colnames(dataset)[1:8]
target_col <- "diabetes"

General Configurations

seed <- 123
if (isTRUE(as.logical(Sys.getenv("_R_CHECK_LIMIT_CORES_")))) {
  # on cran
  ncores <- 2L
} else {
  ncores <- ifelse(
    test = parallel::detectCores() > 4,
    yes = 4L,
    no = ifelse(
      test = parallel::detectCores() < 2L,
      yes = 1L,
      no = parallel::detectCores()
    )
  )
}
options("mlexperiments.bayesian.max_init" = 10L)

Generate Training- and Test Data

data_split <- splitTools::partition(
  y = dataset[, get(target_col)],
  p = c(train = 0.7, test = 0.3),
  type = "stratified",
  seed = seed
)

train_x <- model.matrix(
  ~ -1 + .,
  dataset[data_split$train, .SD, .SDcols = feature_cols]
)
train_y <- dataset[data_split$train, get(target_col)]


test_x <- model.matrix(
  ~ -1 + .,
  dataset[data_split$test, .SD, .SDcols = feature_cols]
)
test_y <- dataset[data_split$test, get(target_col)]

Generate Training Data Folds

fold_list <- splitTools::create_folds(
  y = train_y,
  k = 3,
  type = "stratified",
  seed = seed
)

Experiments

Prepare Experiments

# required learner arguments, not optimized
learner_args <- list(method = "class")

# set arguments for predict function and performance metric,
# required for mlexperiments::MLCrossValidation and
# mlexperiments::MLNestedCV
predict_args <- list(type = "prob")
performance_metric <- metric("auc")
performance_metric_args <- list(positive = "pos")
return_models <- FALSE

# required for grid search and initialization of bayesian optimization
parameter_grid <- expand.grid(
  minsplit = seq(2L, 82L, 10L),
  cp = seq(0.01, 0.1, 0.01),
  maxdepth = seq(2L, 30L, 5L)
)
# reduce to a maximum of 10 rows
if (nrow(parameter_grid) > 10) {
  set.seed(123)
  sample_rows <- sample(seq_len(nrow(parameter_grid)), 10, FALSE)
  parameter_grid <- kdry::mlh_subset(parameter_grid, sample_rows)
}

# required for bayesian optimization
parameter_bounds <- list(
  minsplit = c(2L, 100L),
  cp = c(0.01, 0.1),
  maxdepth = c(2L, 30L)
)
optim_args <- list(
  iters.n = ncores,
  kappa = 3.5,
  acq = "ucb"
)

Hyperparameter Tuning

tuner <- mlexperiments::MLTuneParameters$new(
  learner = LearnerRpart$new(),
  strategy = "grid",
  ncores = ncores,
  seed = seed
)

tuner$parameter_grid <- parameter_grid
tuner$learner_args <- learner_args
tuner$split_type <- "stratified"

tuner$set_data(
  x = train_x,
  y = train_y
)

tuner_results_grid <- tuner$execute(k = 3)
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#> Parameter settings [=======================================================================================>----------] 9/10 ( 90%)
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#> Parameter settings [=================================================================================================] 10/10 (100%)                                                                                                                                    
#>  Classification: using 'classification error rate' as optimization metric.

head(tuner_results_grid)
#>    setting_id metric_optim_mean minsplit   cp maxdepth method
#> 1:          1         0.1860709        2 0.07       22  class
#> 2:          2         0.1860709       32 0.02       27  class
#> 3:          3         0.1860709       72 0.10        7  class
#> 4:          4         0.1860709       32 0.09       27  class
#> 5:          5         0.1860709       52 0.02       12  class
#> 6:          6         0.1860709        2 0.04        7  class

Bayesian Optimization

tuner <- mlexperiments::MLTuneParameters$new(
  learner = LearnerRpart$new(),
  strategy = "bayesian",
  ncores = ncores,
  seed = seed
)

tuner$parameter_grid <- parameter_grid
tuner$parameter_bounds <- parameter_bounds

tuner$learner_args <- learner_args
tuner$optim_args <- optim_args

tuner$split_type <- "stratified"

tuner$set_data(
  x = train_x,
  y = train_y
)

tuner_results_bayesian <- tuner$execute(k = 3)
#> 
#> Registering parallel backend using 4 cores.

head(tuner_results_bayesian)
#>    Epoch setting_id minsplit   cp maxdepth gpUtility acqOptimum inBounds Elapsed      Score metric_optim_mean errorMessage method
#> 1:     0          1        2 0.07       22        NA      FALSE     TRUE   0.044 -0.1860709         0.1860709           NA  class
#> 2:     0          2       32 0.02       27        NA      FALSE     TRUE   0.044 -0.1860709         0.1860709           NA  class
#> 3:     0          3       72 0.10        7        NA      FALSE     TRUE   0.044 -0.1860709         0.1860709           NA  class
#> 4:     0          4       32 0.09       27        NA      FALSE     TRUE   0.044 -0.1860709         0.1860709           NA  class
#> 5:     0          5       52 0.02       12        NA      FALSE     TRUE   0.020 -0.1860709         0.1860709           NA  class
#> 6:     0          6        2 0.04        7        NA      FALSE     TRUE   0.021 -0.1860709         0.1860709           NA  class

k-Fold Cross Validation

validator <- mlexperiments::MLCrossValidation$new(
  learner = LearnerRpart$new(),
  fold_list = fold_list,
  ncores = ncores,
  seed = seed
)

validator$learner_args <- tuner$results$best.setting[-1]

validator$predict_args <- predict_args
validator$performance_metric <- performance_metric
validator$performance_metric_args <- performance_metric_args
validator$return_models <- return_models

validator$set_data(
  x = train_x,
  y = train_y
)

validator_results <- validator$execute()
#> 
#> CV fold: Fold1
#> 
#> CV fold: Fold2
#> 
#> CV fold: Fold3

head(validator_results)
#>     fold performance minsplit   cp maxdepth method
#> 1: Fold1   0.8323638        2 0.07       22  class
#> 2: Fold2   0.7342676        2 0.07       22  class
#> 3: Fold3   0.7959299        2 0.07       22  class

Nested Cross Validation

validator <- mlexperiments::MLNestedCV$new(
  learner = LearnerRpart$new(),
  strategy = "grid",
  fold_list = fold_list,
  k_tuning = 3L,
  ncores = ncores,
  seed = seed
)

validator$parameter_grid <- parameter_grid
validator$learner_args <- learner_args
validator$split_type <- "stratified"

validator$predict_args <- predict_args
validator$performance_metric <- performance_metric
validator$performance_metric_args <- performance_metric_args
validator$return_models <- return_models

validator$set_data(
  x = train_x,
  y = train_y
)

validator_results <- validator$execute()
#> 
#> CV fold: Fold1
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#> CV fold: Fold2
#> CV progress [======================================================================>-----------------------------------] 2/3 ( 67%)
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#> CV fold: Fold3
#> CV progress [==========================================================================================================] 3/3 (100%)
#>                                                                                                                                     
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.
#> 
#>  Classification: using 'classification error rate' as optimization metric.

head(validator_results)
#>     fold performance minsplit   cp maxdepth method
#> 1: Fold1   0.7496034       42 0.02        2  class
#> 2: Fold2   0.6845584       42 0.02        2  class
#> 3: Fold3   0.7959299        2 0.07       22  class

Inner Bayesian Optimization

validator <- mlexperiments::MLNestedCV$new(
  learner = LearnerRpart$new(),
  strategy = "bayesian",
  fold_list = fold_list,
  k_tuning = 3L,
  ncores = ncores,
  seed = seed
)

validator$parameter_grid <- parameter_grid
validator$learner_args <- learner_args
validator$split_type <- "stratified"


validator$parameter_bounds <- parameter_bounds
validator$optim_args <- optim_args

validator$predict_args <- predict_args
validator$performance_metric <- performance_metric
validator$performance_metric_args <- performance_metric_args
validator$return_models <- TRUE

validator$set_data(
  x = train_x,
  y = train_y
)

validator_results <- validator$execute()
#> 
#> CV fold: Fold1
#> 
#> Registering parallel backend using 4 cores.
#> 
#> CV fold: Fold2
#> CV progress [======================================================================>-----------------------------------] 2/3 ( 67%)
#> 
#> Registering parallel backend using 4 cores.
#> 
#> CV fold: Fold3
#> CV progress [==========================================================================================================] 3/3 (100%)
#>                                                                                                                                     
#> Registering parallel backend using 4 cores.

head(validator_results)
#>     fold performance minsplit   cp maxdepth method
#> 1: Fold1   0.7496034       42 0.02        2  class
#> 2: Fold2   0.6845584       42 0.02        2  class
#> 3: Fold3   0.7959299        2 0.07       22  class

Comparison with Logistic Regression

See https://github.com/kapsner/mlexperiments/blob/main/R/learner_glm.R for implementation details.

validator_glm <- mlexperiments::MLCrossValidation$new(
  learner = LearnerGlm$new(),
  fold_list = fold_list,
  ncores = ncores,
  seed = seed
)

validator_glm$learner_args <- list(family = binomial(link = "logit"))
validator_glm$predict_args <- list(type = "response")
validator_glm$performance_metric <- performance_metric
validator_glm$performance_metric_args <- performance_metric_args
validator_glm$return_models <- TRUE

validator_glm$set_data(
  x = train_x,
  y = train_y
)

validator_glm_results <- validator_glm$execute()
#> 
#> CV fold: Fold1
#> Parameter 'ncores' is ignored for learner 'LearnerGlm'.
#> 
#> CV fold: Fold2
#> Parameter 'ncores' is ignored for learner 'LearnerGlm'.
#> 
#> CV fold: Fold3
#> Parameter 'ncores' is ignored for learner 'LearnerGlm'.

head(validator_glm_results)
#>     fold performance
#> 1: Fold1   0.8746695
#> 2: Fold2   0.8751983
#> 3: Fold3   0.8801583

Test Fold Equality

mlexperiments::validate_fold_equality(
  experiments = list(validator, validator_glm)
)
#> 
#> Testing for identical folds in 1 and 2.
#> 
#> Testing for identical folds in 2 and 1.

Predict Outcome in Holdout Test Dataset

preds_rpart <- mlexperiments::predictions(
  object = validator,
  newdata = test_x
)

preds_glm <- mlexperiments::predictions(
  object = validator_glm,
  newdata = test_x
)

Evaluate Performance on Holdout Test Dataset

perf_rpart <- mlexperiments::performance(
  object = validator,
  prediction_results = preds_rpart,
  y_ground_truth = test_y,
  type = "binary"
)

perf_glm <- mlexperiments::performance(
  object = validator_glm,
  prediction_results = preds_glm,
  y_ground_truth = test_y,
  type = "binary"
)
# combine results for plotting
final_results <- rbind(
  cbind(algorithm = "rpart", perf_rpart),
  cbind(algorithm = "glm", perf_glm)
)
# p <- ggpubr::ggdotchart(
#   data = final_results,
#   x = "algorithm",
#   y = "auc",
#   color = "model",
#   rotate = TRUE
# )
# p

Model Comparison